The invention of the gas dynamic CO2 laser and, subsequently, the HF/DF laser, indicated the feasibility of gas-flow lasers of high power. These are continuous wave (cw), supersonic flow devices, and, at a high-power level, have a large saturated gain volume. An optical resonator or cavity is used to resonate photons from the gain volume into a coherent and collimated laser beam. Because the flow is supersonic, the cross-sectional area of this gain volume is roughly rectangular (or trapezoidal) and not a disc. Because of this rectangular gain volume, conventional spherical optical elements in the resonator are difficult to use. If the outcoupled beam from the resonator (the “laser beam”) is rectangular in cross section, it requires clipping and manipulation to transform it into one with a circular cross section. For a laser system, this process is costly and reduces the beam's power. On the other hand, if the beam emanating from the resonator or optical cavity has a circular cross section, then much of the radiative energy inside the rectangular, box-like optical cavity may, unfortunately, contribute to gas heating and/or parasitic lasing. It is, therefore, difficult to couple the radiative energy in a box-like region into an axially-symmetric outcoupled beam.
The first type of optical cavity or resonator that was used for a laser is referred to as a stable resonator and is described in Wrolstad, K. H., Avizonis, P. V. and Holmes, D. A., “Stable Resonators with Increased Fundamental Mode Volume for CO2 Laser Oscillations” J. Phys., Part E, 4, pp. 143-145 (1971). This reference reports obtaining single transverse mode operation (TEM00) with CO2 in a 1.2 cm diameter discharge tube. This beam size diameter, however, is far too small for use with a large, supersonic laser. With a stable resonator, it was realized that operation with a single, low-order mode was not possible. The laser would be multi-mode, which results in an undesirable amount of beam spread when the beam is focused on a distant target. In addition, because the area of the gain volume is rectangular, a stable resonator may fail to fully saturate all of the gain medium. Among other adverse effects, destructive parasitic lasing can then ensue, especially for a large, high-gain laser.
In a conventional chemical oxgen-iodine laser (COIL), the stagnation temperature, static pressure, and Mach number M are about 300 K, a few Torr, and about 2, respectively. The gas is largely helium with most of the balance being oxygen, some of which is in the excited 1Δ electronic state. The stagnation pressure for this gas is about 60 Torr.
Current COIL practice uses a gain medium that is linear rather than toroidal. In the linear case, the mirrors are located at the far end of two optical ducts, away from the supersonic flow where the gain is located. With a purge gas (helium), the ducts are maintained at a slightly higher pressure than the optical cavity to keep the mirrors from degrading. Some means must be provided, such as a material or aerodynamic window, for transmitting the beam into the ambient atmosphere.
All current chemical lasers utilize an unstable resonator for a variety of reasons that include the following important features:                (i) A large mode volume.        (ii) Transverse mode discrimination.        (iii) Single-ended output.        (iv) A confocal beam, if desired.        (v) A central main lobe in the far field.        (vi) The outcoupled beam stems from an annular portion of the coherent wave that is incident on the mirror. Nevertheless, the outcoupled beam is also coherent.See Siegman, A. E., “Stabilizing Output with Unstable Resonator,” Laser Focus, 42, pp. 42-47 (May 1971) for an early discussion of the benefits of this type of resonator. Krupke, W. F. and Sooy, W. R., “Properties of an Unstable Confocal Resonator CO2 Laser System,” IEEE J. Quant. Elect., QE-5(12), pp. 575-586 (1969) describes a three mirror optical system that is current practice. All three mirrors are highly reflective and the resulting beam is toriodal, i.e., in the near field the beam has a hole in it.        
To outcouple a beam, current unstable resonators typically use the diffractive loss from the outer portion of a mirrored surface. In the far field, the hole fills in, but as a result of its annular origin, the side lobes contain a larger fraction of the power than if the beam originated from a uniformly illuminated disc with a diameter equal to the outer annular diameter. FIGS. 1 and 2 illustrate schematically these types of resonators, with the lines having arrows representing the outcoupled beam. In FIG. 1, the outcoupled beam simply diffracts around the smaller mirror. In FIG. 2, a “scraper” mirror is used to outcouple a beam. In both cases, the outcoupled beam in the “near field” (close to the resonator) is annular or donut-shaped. In the “far field” (a distance from the resonator), the beam “fills in,” but does not have an optimal energy distribution.
As mentioned, the saturated gain region is generally box-like in a gas flow laser. There is a misfit between the geometry of this region and an axially symmetric outcoupled beam. In a large laser, this misfit can result in nonuniform saturation of the gain region. A poorly saturated part of the region, means, at least, some loss of laser power, or, at worst, destructive parasitic lasing. On the other hand, more nearly uniform saturation poses the risk of optically coupling gain and absorption regions. The loss of power may be severe when this happens. In this regard, each type of chemical laser is different. For instance, in COIL the absorbing state is the electronic ground state of the iodine atom. This atom is in its diatomic (non-absorbing) form when the fluid is in chemical equilibrium. Because of its low concentration, however, the three-body recombination rate of iodine atoms is slow. Hence, ground state iodine atoms can remove energy from a coherent radiative field, and then lose this energy by spontaneous emission or collisional deactivation. This loss process can occur in the absorbing, supersonic flow just downstream of the saturated gain region.
Accordingly, a need exists for a resonator or optical cavity design that is particularly adapted to the modern, supersonic flow chemical laser, in particular to the COIL.